Antiferromagnetic Spin Fluctuation above the Superconducting Dome and the Full-Gaps Superconducting State in LaFeAsOi^F,,, Revealed by 75 As-Nuclear Quadrupole Resonance (N o o o =5 a- -a o (N > CN O X T. Oka 1 , Z. Li 2 , S. Kawasaki 1 , G. F. Chen 2 , N. L. Wang 2 , and Guo-qing Zheng 1 ^ 2 1 Department of Physics, Okayama University, Okayama 700-8530, Japan and 2 Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100190, China (Dated: January 30, 2012) We report a systematic study by 75 As nuclear-quadrupole resonance in LaFeAsOi-^Fa;. The antiferromagnetic spin fluctuation (AFSF) found above the magnetic ordering temperature Tn — 58 K for x — 0.03 persists in the regime 0.04 < x < 0.08 where superconductivity sets in. A dome- shaped ^-dependence of the superconducting transition temperature T c is found, with the highest T c = 27 K at x — 0.06 which is realized under significant AFSF. With increasing x further, the AFSF decreases, and so does T c . These features resemble closely the cuprates L^-^Sr^CuCU. In x — 0.06, the spin-lattice relaxation rate (1/Ti) below T c decreases exponentially down to 0.13 T c , which unambiguously indicates that the energy gaps are fully-opened. The temperature variation of 1/Ti below T c is rendered nonexponential for other x by impurity scattering. The discovery of superconductivity in LaFeAsOi-^Fj, at the transition temperature T c = 26 K [l| has gained much attention in the condensed-matter physics commu- nity. The electron -doping (F -doping) suppresses the antiferromagnetic ordering at TV — 140 K in LaFeAsO and high-T c superconductivity appears [lj. The T c sig- nificantly increases up to 55 K in RFeAsOi_ x F x (R: Ce, Pr, Nd, Sm) [2j, |3j]. To elucidate the mechanism of Cooper pairs formation in these arsenides, it is essen- tial to know the superconducting gap symmetry and the normal-state properties. Previous nuclear-magnetic reso- nance (NMR) and nuclear-quadrupole resonance (NQR) measurements have found that the superconductivity is in the spin-singlet state with multiple gaps 4-6]. Re- cent systematic measurements on Ba(Fei_ 2: Co 2 ,) i 2As2[7|, CaFe2As 2 under pressure 13, LaNiAsOi-^F^^], and BaFe2(Asi_ x P a: )2[lO| have suggested that the antiferro- magnetic spin fluctuation (AFSF) originated from their multiple electronic bands correlates with the appearance of the pertinent superconducting properties. On the other hand, there are also reports suggesting that AFSF is not important to realize high T c . 11 1 For prototypical LaFeAsOi-^F^, several issues remain elusive. One is the role of AFSF. In cuprates, it has been believed that AFSF plays a crucial role to induce high-T c superconductivity, but the situation in LaFeAsOi^F^ is still unclear [5j, Il2|- 161. Some previous studies by NMR found no AFSF [UO. The second issue is the doping dependence of T c . It was initially reported that T c forms a wide plateau at 0.04 < x < 0.12 [l|, which raises a question about the effect of doping. The third unresolved issue is the super- conducting gap symmetry. The spin-lattice relaxation rate (1/Ti) decreases sharply below T c , but the data were insufficient for dis ting uishing between d-wave from sign- reversal s-wave |5j, ll2Hl6j ] . From other experimen- tal probes, some measurements suggested the existence of node [Tjj , but the photoemission spectroscopy and the point contact Andreev reflection measurement suggested a nodeless gap 18[ [l9j . Here we report results of systematic 75 As NQR stud- ies on LaFeAsOi-^F,, (x = 0.03, 0.04, 0.06, 0.08, 0.10, and 0.15). An antiferromagnetic order with TV = 58 K is found for x — 0.03. Bulk superconductivity sets in at T c = 21 K for x = 0.04, with strong AFSF. A dome-shaped x-dependence of T c is found, with the highest T c = 27 K at x — 0.06 which is realized un- der significant AFSF. With further doping, the AFSF is weakened and disappears for x > 0.10. Concomi- tantly, T c decreases. These features resemble closely the case of cuprates I^-^Sr^CuO^ and suggests that the AFSF is important in producing the superconductivity in LaFeAsOi-zFa; as well. The systematic observation of the AFSF in the low-doping regime is unprecedented, and the high quality samples enable us to reveal a dome shape of the T c which has a maximum at quite low x. In the superconducting state, 1/Ti for x = 0.06 decreases exponentially down to 0.13 T c , which is clear and direct evidence for a fully -gapped superconducting state. The T- variation of 1/Ti below T c is rendered nonexponential for x either smaller or larger than 0.06, showing a seem- ingly T 3 behavior for x — 0.10, which is accounted for by impurity scattering. The polycrystalline samples of LaFeAsOi^^F^ (x — 0.03, 0.04, 0.06, 0.08, 0.10, and 0.15) were synthesized by the solid state reaction method |20l.l2l|. Here, x indi- cates the nominal composition of the starting material. Quite often, resistivity measurements give a higher T c than magnetic susceptibility or NQR. We define T c by the latter methods, ac susceptibility measurements using the in-situ NQR coil indicate T c = 21, 27, 23, 18, and 12 K for x = 0.04, 0.06, 0.08, 0.10, and 0.15, respectively. The 1/Ti decreases exactly below such-determined T c . The Ti is determined by an excellent fit to the single exponential curve 1 — ^jp- — exp(^^) 21|, where M MM - m - and M(t) are the nuclear magnetization in the thermal equilibrium and at a time t after the saturating pulse, respectively. right above Tjv and then decreases below, leaving a tiny peak at T/v- For x > 0.04, 1/Ti decreases rapidly be- low T c due to the opening of the superconducting energy gaps. Z 10 12 14 Frequency (MHz) 0.04 0.08 0.12 F content x FIG. 1: (color online) (a) Doping dependence of the 75 As- NQR spectrum for LaFeAsOi-^Fa; measured above T c . Data for x — 0.08 are from Ref.[g|. Solid curves are Lorentzian fittings which give a FWHM of ~ 0.95, 1.2, 1.8, and 2.3 MHz for x = 0.06, 0.08, 0.10, and 0.15, respectively, (b) The spec- tra above and below TV — 58 K for x — 0.03. (c) The x dependence of vq . Figure 1 (a) shows the 75 As-NQR spectrum for 0.04 < x < 0.15 measured above T c . As seen in the fig- ure, a clear single peak, which can be fitted by a single Lorentzian curve, is observed for x > 0.06. The spectra do not change below T c . However, we observed two peaks for x = 0.03 (Fig. 1(b)) and 0.04. This indicates that there are two As -sites which are in different surround- ings. The NQR frequency vq increases with increasing x, as seen in Fig. 1(c). Here, ^q probes the electric- field gradient generated by the carrier distribution and the lattice contribution surrounding the As nucleus. The doping evolution of vq, the spectral shapes, and the sin- gle component of T\ indicate that the electron carriers were homogeneously doped for x > 0.06, but phase sep- aration occurs in x — 0.03 and 0.04. We speculate that this may be due to the local distribution of the F ion around the As nucleus, which is inevitable in a quite low- doping region. Remarkably, the T dependences of \jT\ measured at each peak of X = 0.03 and 0.04 indicate that each phase is homogeneous. The same behavior was found in Rcf.[22], where the vq is quite similar to ours alth oug h the nominal x there is larger than ours by ~ 0.02 [ilj. Figure 1(b) shows the T evolution of the NQR spectra for x = 0.03. Below TV = 58 K, the spectra are broadened due to an antiferromagnetic order. Figures 2(a) and 2(b) show the T dependence of 1/Ti for all samples. For x = 0.03, 1/Ti shows a small upturn 0.0001 'As-NQR a x = 0.03 V x = 0.04(L) o x = 0.04(H) • x = 0.06 (b) 10 100 T(K) 10 100 T(K) FIG. 2: (color online) The T dependences of 1/Ti for x = 0.03, 0.04 and 0.06 (a) and for x = 0.08, 0.10, and 0.15 (b). Data for x = 0.03 were collected at the high-frequency (H) NQR peaks. For x — 0.04, 1/Ti was measured at both the low- frequency (L) and the H peak. Solid curves below T c for x > 0.04 are the simulations based on a s wave superconducting gap model with impurity scattering (see the text). The dashed line indicates the relation 1/Ti cc T 3 . The dotted and solid arrows indicate Tn and T c , respectively. Before going into the detail of the superconducting state, we first discuss the normal-state property. For this purpose, we plot \jT\T vs T in Fig. 3. None of the samples shows a Korringa relation \jT\T = const, expected for a conventional metal. Above Tn of x = 0.03, 1/TiT increases with decreasing T due to the AFSF. Such AFSF persists in x = 0.04, 0.06, and 0.08, where \jT{T increases with decreasing T down to T c . To model the 1/TiT above Tjy or T c , we employed the th eory for a weakly antiferromagnetically-correlated metal |23j j , l/^T =(1/T 1 T) AF +(1/T 1 T) = C/(T + 9) + (\/T x T) . Here, the first term described the contribution from the antiferromagnetic wave vector, and the second term is the contribution from the density of states (DOS) at the Fermi level. For x = 0.03, 9 is simply — TV, where the data can be well fitted except around Tjv [24|. As seen in Fig. 3, 1/TiT for x = 0.04, 0.06, and 0.08 are well 'o U' i ■ i ■ • x = 0.06 Ax = 0.08 * ». ♦ x = 0.10 4 s A ^S * D x = 0.15 ♦ ♦ ♦ n 1 .< D D n n n W* , i = 0.03 ).04(L) 1.04(H) fi * ~% [5* B V fi A 250 300 FIG. 3: (color online) T dependence of \/T\T for various x. The curves above Tn or T c are fits to the AFSF theory (see the text). The inset is the enlarged part for 0.06 < x < 0.15. reproduced by this model with 9 ~ 10, 25 and 39 K, respectively. The low-frequency NQR peak for x — 0.04 gives a smaller fl~5K. The increase of 9 with increasing x means that the system moves away from the magnetic instability (MI) where 9 = K. With further doping, for x = 0.10 and 0.15, no enhancement of \jT\T is seen. Instead, \jT\T decreases with decreasing T, which was recently explained by the loss of the DOS due to a topo- logical change of the Fermi surface [9j, |25j . The results in previous reports of the lack of the AFSF for x > 0.10 12J, [TJ, LL5J are consistent with our results for x = 0.10 and 0.15. The remarkable finding is that the highest T c — 27 K is realized at x = 0.06, which is away from the MI. This situ- ation is quite similar to the cuprates La2- K Sr :r Cu04 26]. In the scenario of spin fluctuation-mediated supercon- ductivity, this can be understood as follows. At high doping levels, the decrease of T c is due to the weaken- ing of the AFSF. In the vicinity of the MI, on the other hand, the too strong low-energy fluctuation acts as pair breaking|27]. Therefore, a maximal T c is realized at some point away from the MI with moderate AFSF. Figure 4 shows the phase diagram for LaFeAsOi-^F-r obtained in the present study. The most important find- ing is that the highest T c is found in the low-doping regime, which makes our T c vs x relation look like a dome shape. In the previous study [l|, [28|, the fail- ure of obtaining higher T c in the low-doping regime is probably due to sample inhomogeneity as evidenced by the broader (in fact, two-peak- featured) NQR spec- trum [29:] . The present phase diagram is consistent with that for Ba(Fei_ 2; Coi C )2As2[7| but is somewhat different from that for BaFe? (Asi -^P^)? [10] . whose T c shows a maximum around 9 — 0. This slight difference may originate from the difference of the tuning parameter for their ground states. The ground states for both LaFeAsOi-^Fa; and Ba(Fei_^Co z ,)2As2 are tuned by electron doping. On the other hand, isovalent P dop- ing acts as chemical pressure on BaFe2(Asi_ a; P 2 ,)2. In any case, these phase diagrams support the intimate re- lationship between AFSF and superconductivity in iron arsenides. Furthermore, such a phase diagram has consis- tently been found in high-T c cuprate La 2 - 2; Sr a .Cu04 26] and heavy fermion compounds [30j, indicating that the AFSF plays a significant role to induce superconductiv- ity in strongly correlated electron systems in general. (a) Q5 (b) ^ 45 30 15 60 a in <U O, 30 £ 15 - / - / / / / / ■ / -A T N LaFeAsOi -X-Tx AF ■ • T • • c ■ - sc • 4 0.02 0.04 0.06 0.08 0.10 0.12 0.14 F-content (x) FIG. 4: (color online) Phase diagram obtained in this study. AF and SC denote the antiferromagnetically ordered and su- perconducting states, respectively, (a) x dependence of 9. The dotted line is a guide to the eyes. The shade indicates the region of phase separation, (b) x dependence of Tn and T c determined by NQR measurements. Next, we turn to the superconducting state. Figure 5(a) shows the T dependence of \jT\ for x = 0.06. Be- low T c , 1/Ti decreases steeply due to the opening of the superconducting gaps. The hump structure at T ~ 0.4 T c is due to the multiple-gap character as reported for other compounds [4j-|6j, [8(. The T variation at low T is much stronger than T 3 , and even stronger than T 5 , as can be clearly seen in the figure. In fact, 1/Ti decreases exponentially below 0.4 T c . In Fig. 5(b), we plotted 1/Ti against T c /T in a semilogarithmic scale. As indicated by the solid line, the 1/Ti below T ~ 0.4 T c clearly fol- lows the relation 1/Ti oc exp(— A /fcsT) with Aq/ ksT c = 1.8, where A and fcs denote the gap size at T = and the Boltzmann constant, respectively. This is clear and direct evidence that the superconducting state is fully gapped in LaFeAsO0.94F0.06- ( a ) 10 0.1 0.01 0.001 0.0001 5 As-NQR 10 100 T(K) FIG. 5: (color online) (a) The T dependence of 1/Ti for x = 0.06. (b) Semilogarithmic plot of 1/Ti vs T c /T. The solid line represents the relation 1/Ti oc exp(— — j0- The evolution of the superconducting-state properties can be seen in Fig. 2. For x — 0.06-0.10, 1/Ti shows a marked hump structure around T ~ 0.4 T c and is followed by a still sharper decrease below. However, the low-T behavior of 1/T) changes gradually, as to decrease less and less steeply as x increases. Eventually, for x = 0.15, the hump structure disappears completely. Instead, a simple T dependence emerges which is close to T 3 . Such T 3 behavior has been reported previously |12h14 and was taken as evidence for line nodes. Below we show that it is a consequence of impurity scattering. Namely, the T 3 is an accidental one rather than an intrinsic one. In fact, in Bai_ x K x Fe2As2, the low-T behavior of 1/Ti also changes when the sample purity differs Assuming sign reversing s-wave symmetry impurity scattering, one can reproduce the evolution of the 1/Ti below T c . By introducing the impurity scatter- ing parameter r\ in the energy spectrum in the form of E = uj+ii], the 1/Ti in the superconducting state is given with by Ti(T c ) ~~ 4T J- Ti(T) ' T 4T W GG = [(Re{(w ^%(VF GG + Wff) [3| where + iT,)/J(u> + ir,)* + \A(k F )\*}) kw } 2 and W FF = [(Re{l/v/(a; + irtf + |A(A; F )| 2 }A(fc F )) feF ] 2 . Here the A is the gap parameter, and (. . .) is the average over the entire Fermi surface, and runs over three bands consisting of two hole pockets at the T point and an electron pocket at the M point, respec- TABLE I: The fitting parameters AJ" ( = A 3 ), A 2 , r\ in the unit of k B T c , and Ni:N 2 :N 3 . X Tc(K) A+ A- n Nv.Nr.Na 0.04(L) 21 4.50 0.93 0.39 0.335:0.330:0.335 0.04(H) 21 4.58 1.63 0.27 0.38:0.24:0.38 0.06 27 5.62 1.11 0.006 0.30:0.40:0.30 0.08 23 3.37 0.92 0.03 0.303:0.394:0.303 0.10 18 3.00 0.83 0.035 0.305:0.39:0.305 0.15 12 2.62 0.79 0.15 0.31:0.38:0.31 tively [3S]. Namely, for a quantity F, (T [A(k F )]) kp = [7ViT(A+) + N 2 F{Az) + iV 3 F(Aa )] /(JVi + N 2 + N 3 ), where Ni is the DOS coming from band i (i = 1,2,3). Here, it is tempting to assign bands 1, 2, and 3 to the 7, /3, and a bands found in angle-resolved photoemission spec- troscopy measurement [36j. It is noted that the weaker T dependence in the x = 0.15 sample can be understood as due to the impurity scattering that brings about a fi- nite DOS. For x = 0.04 where two As sites were found, 1/Ti for each site can also be fitted by the same model, with an additional feature that a large 77 is needed to ex- plain the low-T behavior. This can be understood if the two phases coexist in the nanoscale [22| , where one phase acts as an impurity scatterer for the other. The obtained fitting parameters are summarized in Table 1. Finally, we note that an s ++ wave [37f seems difficult to explain the lack of the coherence peak just below T c and the x evolution of low-T behavior of 1/Ti. In conclusion, we have presented the results of sys- tematic NQR measurements on high quality samples of LaFeAsOi-zF^ (x = 0.03, 0.04, 0.06, 0.08, 0.10, and 0.15). The AFSF seen above T N = 58 K of x = 0.03 persists in the 0.04 < x < 0.08 regime. The highest T c = 27 K is realized for x — 0.06 which is away from the magnetic instability but with significant AFSF. The phase diagram closely resembles those of the cuprates La2_ 3; Sr 3 ;Cu04 and other iron arsenides, which suggests that the AFSF is also important to produce the super- conductivity in LaFeAsOi-ajFj,. In x = 0.06, 1/Ti below T c decreases exponentially down to 0.13 T c , which un- ambiguously indicates that the superconducting gaps are fully-opened. The T-variation of 1/Ti below T c is ren- dered nonexponential for x either smaller or larger than 0.06, which is accounted for by impurity scattering. We thank M. Ichioka for help in the calculation and K. Ishida for useful communication. Work in Okayama was supported in part by research grants from MEXT (No. 22103004 and 23102717). Work in Beijing was supported by CAS and NSFC. [1] Y. Kamihara, T. Watanabe, M. Hirano, and H. Hosono, J. Am. Chem. Soc. 130, 3296 (2008). [2] X. H. Chen, T. Wu, G. Wu, R. H. Liu, H. Chen, and D. F. Fang, Nature 453, 761 (2008). [3] Z.-A. Ren, W. Lu, J. Yang, W. Yi, X.-L. Shen, Z.-C. Li, G.-C. Che, X.-L. Dong, L.-L. Sun, F. Zhou, and Z.-X. Zhao, Chin. Phys. Lett. 25, 2215 (2008). [4] K. Matano, Z. A. Ren, X. L. Dong, L. L. Sun, Z. X. Zhao, and G. -q. Zheng, Europhys. Lett. 83, 57001 (2008). [5] S. Kawasaki, K. Shimada, G. F. Chen, J. L. Luo, N. L. Wang, and G. -q. Zheng, Phys. Rev. B 78, 220506(R) (2008). [6] K. Matano, Z. Li, G. L. Sun, D. L. Sun, C. T. Lin, M. Ichioka, and G. -q. Zheng, Europhys. Lett. 87, 27012 (2009). [7] F. L. Ning, K. Ahilan, T. Imai, A. S. Sefat, M. A. McGuire, B. C. Sales, D. Mandrus, P. Cheng, B. Shen, and H.-H Wen, Phys. Rev. Lett. 104, 037001 (2010). [8] S. Kawasaki, T. Tabuchi, X. F. Wang, X. H. Chen, and G. -q. Zheng, Supercond. Sci. Tech. 23, 054004 (2010). [9] T. Tabuchi, Z. Li, T. Oka, G. F. Chen, S. Kawasaki, J. L. Luo, N. L. Wang, and G. -q. Zheng, Phys. Rev. B 81, 140509(R) (2010). [10] Y. Nakai, T. lye, S. Kitagawa, K. Ishida, H. Ikeda, S. Kasahara, H. Shishido, T. Shibauchi, Y. Matsuda, and T. Terashima, Phys. Rev. Lett. 105, 107003 (2010). [11] H. Kinouchi, H. Mukuda, M. Yashima, Y. Kitaoka, P. M. Shirage, H. Eisaki, and A. Iyo, Phys. Rev. Lett. 107, 047002 (2011). [12] Y. Nakai, K. Ishida, Y. Kamihara, M. Hirano, and H. Hosono, J. Phys. Soc. Jpn. 77, 073701 (2008). [13] H.-J. Grafe, D. Paar, G. Lang, N. J. Curro, G. Behr, J. Werner, J. Hamann-Borrero, C. Hess, N. Leps, R. Klingeler, and B. Buchner, Phys. Rev. Lett. 101, 047003 (2008). [14] H. Mukuda, N. Terasaki, H. Kinouchi, M. Yashima, Y. Kitaoka, S. Suzuki, S. Miyasaka, S. Tajima, K. Miyazawa, P. Shirage, H. Kito, H. Eisaki, and A. Iyo, J. Phys. Soc. Jpn. 77, 093704 (2008). [15] Y. Nakai, S. Kitagawa, K. Ishida, Y. Kamihara, M. Hi- rano, and H. Hosono, New. J. Phys. 11, 045004 (2009). [16] Y. Kobayashi, E. Satomi, S. C. Lee, and M. Sato, J. Phys. Soc. Jpn. 79, 093709 (2010). [17] C. Martin, M. E. Tillman, H. Kim, M. A. Tanatar, S. K. Kim, A. Kreyssig, R. T. Gordon, M. D. Vannette, S. Nandi, V. G. Kogan, S. L. Budko, P. C. Canfield, A. I. Goldman, and R. Prozorov, Phys. Rev. Lett. 102, 247002 (2009). [18] T. Sato, S. Souma, K. Nakayama, K. Terashima, K. Sug- awara, T. Takahashi, Y. Kamihara, M. Hirano, and H. Hosono, J. Phys. Soc. Jpn. 77, 063708 (2008). [19] R. S. Gonnelli, D. Daghero, M. Tortello, G. A. Um- [20 [21 [22 [23 [24 [25 [26 [27 [28 [29 [30 [31 [32 [3:5 [34 [35 [36 [37] marino, V. A. Stepanov, R. K. Kremer, J. S. Kim, N. D. Zhigadlo, J. Karpinski, Physica C 469, 512 (2009). G. F. Chen, Z. Li, G. Li, J. Zhou, D. Wu, J. Dong, W. Z. Hu, P. Zheng, Z. J. Chen, H. Q. Yuan, J. Singleton, J. L. Luo, and N. L. Wang, Phys. Rev. Lett. 101, 057007 (2008). See Supplemental Material at |http://link.aps.org/supplemental/10.1103/PhysRevLett.l08.047001 for details on sample quality and x dependence of vq . G. Lang, H.-J. Grafe, D. Paar, F. Hammerath, K. Man- they, G. Behr, J. Werner, and B. Buchner, Phys. Rev. Lett. 104, 097001 (2010). T. Moriya, Spin fluctuations in itinerant magnetism, Springer, 1985, Berlin. The deviation may be understood as due to impurity scattering by another phase that coexists in the nanoscale m. H. Ikeda, R. Arita, and J. Kunes, Phys. Rev. B 82, 024508 (2010). S. Ohsugi, Y. Kitaoka, K. Ishida, G. -q. Zheng, and K. Asayama, J. Phys. Soc. Jpn. 63, 700 (1994). T. Moriya and K. Ueda, J. Phys. Soc. Jpn. 63, 1871 (1994). H. Luetkens, H.-H. Klauss, M. Kraken, F. J. Litterst, T. Dellmann, R. Klingeler, C. Hess, R. Khasanov, A. Amato, C. Baines, M. Kosmala, O. J. Schumann, M. Braden, J. Hamann-Borrero, N. Leps, A. Kondrat, G. Behr, J. Werner, and B. Buchner, Nature Mater. 8, 305 (2009). S. Kitagawa, Y. Nakai, T. lye, K. Ishida, Y. Kamihara, M. Hirano, and H. Hosono, Physica C 470, S282 (2010). N. D. Mathur, F. M. Grosche, S. R. Julian, I. R. Walker, D. M. Freye, R. K. W. Haselwimmer, and G. G. Lon- zarich, Nature 394, 39 (1998). Z. Li, D. L. Sun, C. T. Lin, Y. H. Su, J. P. Hu, and G. -q. Zheng, Phys. Rev. B 83, 140506(R) (2011) . I. I. Mazin, D. J. Singh, M. D. Johannes, and M. H. Du, Phys. Rev. Lett. 101, 057003 (2008). K. Kuroki, S. Onari, R. Arita, H. Usui, Y. Tanaka, H. Kontani, and H. Aoki, Phys. Rev. Lett. 101, 087004 (2008). Z. Li, Y. Ooe, X. C. Wang, Q. Q. Liu, C. Q. Jin, M. Ichioka, and G. -q. Zheng, J. Phys. Soc. Jpn. 79, 083702 (2010). D. J. Singh and M. H. Du, Phys. Rev. Lett. 100, 237003 (2008). H. Ding, P. Richard, K. Nakayama, K. Sugawara, T. Arakane, Y. Sekiba, A. Takayama, S. Souma, T. Sato, T. Takahashi, Z. Wang, X. Dai, Z. Fang, G. F. Chen, J. L. Luo, and N. L. Wang, Europhys. Lett. 83, 47001 (2008). H. Kontani and S. Onari, Phys. Rev. Lett. 104, 157001 (2010).